There was a recent question on the OR-Tools forum that asked how to minimize a passenger’s in-vehicle travel time. I summarized what I had done in the past in my previous entry. The OP replied in the final post in the thread that Mizux’s suggestion to use SetPathEnergyCostOfVehicle() was the best solution. I’ve never used that before, so in this post I am revisiting my previous post using that new function.

There was a recent question on the OR-Tools forum that asked how to minimize a passenger’s in-vehicle travel time. I’ve done this before, but it has been a while, so this post summarizes what I do in similar problems.

A question came up a last fall on the OR-Tools forum about how to enforce a chained pickup and delivery problem. The standard pickup and delivery problem is to pick something up from one place, and then deliver that thing to some other place. This question asked about an interesting and fairly common variation of the standard PDP: a chained sequence of pickups and deliveries.

In my previous post, I looked at how slack variables were used to determine the maximum cost value of a node on a route, and how that value could be assigned as a “cost” for the entire route. The program, vrp_node_max.py. was developed by Mizux and is really clever. However, as I mentioned in my previous post, the final note in the original forum thread from Guy was a note that the approach did not seem to work in practice. With this post, I want to explore what that means and how to fix it.

A few weeks ago while working on a project I came across a very interesting program in the OR-Tools code base. The origin of the code started from a question from Guy on the OR-Tools forum. When the question was originally posted last summer, I didn’t have the opportunity to dig into the proposed result, but now I do, and I discovered it is a very interesting (ab)use of the slack dimension. Since I haven’t written anything here in ages, it seemed like a good idea to write up a blog post on the code and my understanding of how it works.

Recently a question was asked on the OR Tools forum about how to convert a single-day TSP problem into a multi-day TSP problem. The original poster was thorough, and cross posted to Stack Overflow and posted his original code as a gist on GitHub. The problem is interesting in that it takes a away from an idealized lab problem toward integrating real-world constraints. So I responded to the question with my solution, and decided to write it up properly here.

A lot of stuff in the ClojureScript/React world is documented through implication, as in “In a similar way you can do X, Y, and Z just as you would in a React project without much difficulty”. But I don’t know React, so I run into difficulties and therefore need explicit examples. This post is what I wish I’d found when I originally searched for “material design icons and clojurescript”, documenting how I integrated Material Design Icons into a Helix-based ClojureScript app.This is a note-to-self style blog post. I am building a new website using Postgraphile and ClojureScript. On the ClojureScript side of things, I’m using helix for the moment, seeing how far local data inside of “modern React development” will get me. Today I managed to integrate Material Design Icons, and it took about a half-hour to figure out the semantics, so I want to document what I did.

My last post on this blog (it’s been a while, apparently) documented how I set up Traefik to front my Elixir and ClojureScript Docker containers. In that post I said that I didn’t yet know how to expose two ports from a container. Well, yesterday I needed to do that, so I searched the internet and figured it out, and I’m writing it up here to cross off that TODO item.

I figured out how to set up the Traefik Edge Router to front my Elixir and ClojureScript Docker containers. My setup is a bit strange, and I couldn’t find a clear example in the extensive Traefik docs. So I’m writing this up in a quick post.

Continuing the theme of writing up my work on optimizing street sweeping routes for Glendale, California, in this post I’m going to talk about how I made an animation to visualize the solutions generated by Google’s OR-Tools.